nCight | nCight, Inc: Building the next generation of physician networks through data unions. | Round 16

Project Name


Project Description

Creation of a prediction algorithm that determines whether an arthroscopic image is classified as a knee or shoulder.

Final Product

not set

Proposal One Liner

Creating a platform for physician engagement on the Ocean platform via use case with arthroscopic image labeling.

Proposal Description

Independent physicians struggle with maintaining autonomy due to the negative reimbursement pressures from third party payors. For the first time in the history of the United States, the number of employed physicians outnumber independent physicians. Orthopedic surgeons have been able to withstand the consolidating pressures largely due to access to ancillary streams of revenue such as ambulatory surgery centers and owning PT and advanced imaging (e.g. MRI). Even with these additional streams of income the number of surgeons in private practice has declined by over 30% over the last 20 years. Independent physicians drive innovation and represent an important part of the healthcare system. It was independent surgeons that pioneered advances in pain management and anesthesia to make outpatient procedures safe and cost effective, saving the healthcare system 100’s of millions of dollars. There is therefore a public interest in keeping physicians independent. By focusing on this group of physicians with a strong entrepreneurial spirit and fierce desire to maintain autonomy we believe this represents an early adopter group. We aim to engage independent Orthopedic Surgeons into the data economy by leveraging the Ocean Protocol to turn data assets in the form of surgical images into tokenized assets that will be provided to medical device data consumers. The orthopedic medical device market is a $39.5 B industry. Limited access to surgeon use data makes market analysis and servicing of current products difficult. In comparison, the $2.6T pharmaceutical industry has access to granular data through companies like Iqvia ($48B) that assist marketing and decision making, there is no equivalent in the medical device space. This is where our opportunity lies to provide granular, unique data to a specific customer segment from data producers motivated to engage. nCight, Inc utilizing the Ocean protocol will establish a network of orthopedic surgeons and their patients to align incentives of patients, physicians and data consumers creating a marketplace for data assets. nCight, Inc will function as a trusted party through relationships within the orthopedic surgery world to bootstrap the network. We will be creating a business to business marketplace.The initial approach will be to consolidate one side of this market by building our physician network of data assets. Care will be taken to not build something with no buyers so we will also conduct experiments testing our value hypothesis that medical device companies will engage with nCight, Inc if we build a regional network of physicians that demonstrate our ability to scale and engage with a segment of physicians. nCight, Inc will serve as a service intermediary connecting the crypto immature physician network to crypto immature medical device companies. We anticipate as adoption and comfort increases, more participants will manage these data relationships internally.

Proposal Goals

1. Develop a mobile app UI/UX prototype that shows capabilities of data union for physician contributors and labelers of arthroscopic surgical images. (COMPLETED)

2. Engage data scientists to develop an accurate prediction model for determining which surgical site an arthroscopic image should be classified into (knee or shoulder). (IN PROGRESS)

3. Develop mobile app allowing physician subject matter experts to engage with Ocean protocol through gamification and rewards to label arthroscopic images. (IN PROGRESS)

4. Create platform that unlocks healthcare data

We will provide access to a private dataset and incentive data scientists to participate in development of a prediction algorithm. We will be narrowly focused on simply determining if an image should be classified as a knee or shoulder image. This classification capability will serve as a springboard for enabling physicians to utilize de-identified, HIPPA compliant data and extract value. In the first portion of our project we were able to create a mockup of the surgeon UX to encourage physician participation. In this round of the project we will deploy a private dataset and train an algorithm and attempt to achieve a high accuracy level for determining the surgical site of arthroscopic images.

We have already piloted and deployed on the ocean marketplace a sample dataset contract.

Additionally, we have deployed a testnet dataset for testing of the knee/shoulder dataset.

Grant Deliverables

  • Grant Deliverable 1: Deployment of private compute to data dataset on the Ocean Market Place (test network)
  • Grant Deliverable 2: Creation of prediction algorithm, working with Algovera team deployed on Ocean test network
  • Grant Deliverable 3: Results of data scientist competition to run additional algorithms against dataset (published on Mainnet)
  • Grant Deliverable 4: Mobile App that enables creation of data union for sharing of arthroscopic images, labeling of images and reward allocation.

Value Add Criteria

  1. Usage - we believe healthcare is a data market that is quite immature but there is great need within the field.  The price discovery, and privacy associated with the AMM tools and Compute to Data are an attractive entryway into the data economy in a field with significant security concerns and regulations.  We anticipate with growth of a physician network the # of assets and total value locked within the ocean protocol should grow as a function of physician referrals.
  2. Viability -  Our team has significant healthcare experience and has the advantage of understanding the market of the users that would be interested in participating in an activity that helps to maintain physician autonomy. We believe we are uniquely situated to execute on the plan as laid out.

Funding Requested


Wallet Address


1 Like

Progress has been made on some of the deliverables to be delivered and a portion of this grant is therefore retroactive in nature.

  1. The algorithm for knee/shoulder prediction has been deployed on the Ocean marketplace. Please see tutorial link: Tutorial for using token-gated apps on HuggingFace | Algovera Docs
  2. There is also a video supplement to this tutorial: HuggingFace & Ocean Protocol App Demo - YouTube
  3. The sample dataset:
    4.Sample algorithm: